11 月 20 日，华为乾崑生态大会在广州盛大启幕，会上鸿蒙座舱携 MoLA 混合大模型 Agent 架构重磅亮相，不仅首推座舱智能化 L1 - L5 五级分级标 准，更以全感官体验、自然交互与主动服务三大方向，重塑智能座舱行业发展格局。目前鸿蒙座舱用户数已突破 120 万，2025 年 1 - 10 月问界车型搭载版 本的平均净推荐值达 93.6%，成用户口碑之选。 作为 HarmonySpace 5 起采用的核心架构，MoLA 通过融合多模态感知、通用大模型与垂域智能体，搭建起从用户意图到硬件执行的完整链路。其独创 的全域免唤醒、导航三模式纠错等功能，让语音交互告别指令束缚。用户说 "车里有点暗" 即可开启阅读灯，导航时可随时改口调整目的地，彻底实现口语 化控车。 此次大会明确鸿蒙座舱率先进入 L3 专业助理阶段。相较于前两级，L3 阶段具备长期记忆、模糊意图理解等核心能力，可主动提供服务。依托感知力、 记忆力等六大类人能力，座舱能记住用户通勤习惯并主动创建场景，还可通过图文视频混合形式解答用车疑问，甚至联动外部工具完成咖啡预订等跨场景任 务。 华为通过行业 Know - How、全栈 AI 算法等能力， ...

据介绍，相较于前两级，L3 阶段具备长期记忆、模糊意图理解等核心能力，可主动提供服务。依托感 知力、记忆力等六大类人能力，座舱能记住用户通勤习惯并主动创建场景，还可通过图文视频混合形式 解答用车疑问，甚至联动外部工具完成咖啡预订等跨场景任务。 责任编辑：杨赐 会上，鸿蒙座舱携 MoLA 混合大模型 Agent 架构亮相。作为 HarmonySpace 5 起采用的核心架构， MoLA 通过融合多模态感知、通用大模型与垂域智能体，搭建起从用户意图到硬件执行的完整链路。 新浪科技讯 11 月 21日上午消息，在华为乾崑生态大会上，华为智能汽车解决方案 BU CEO 靳玉志发表 演讲。靳玉志表示，鸿蒙座舱率先进入 L3 专业助理阶段。 官方数据显示，目前鸿蒙座舱用户数已突破 120 万，2025 年 1 - 10 月问界车型搭载版本的平均净推荐 值达 93.6%。 ...

Core Insights - Huawei's autonomous driving solution, QianKun Intelligent Driving (ADS), has achieved significant milestones, including 1 million units of delivery and 40 billion kilometers of assisted driving mileage [1][5][9] Group 1: ADS Development and Features - The ADS has undergone three commercial iterations, with ADS 1 launched in 2022 for urban NCA, ADS 2 in 2023 enabling nationwide driving without high-precision maps, and ADS 3 offering end-to-end and parking functionalities [3][5] - The recently released ADS 4 has upgraded its technical architecture and user experience, providing comprehensive support for various driving scenarios, including parking and emergency braking [5][6] - As of July 2023, Huawei's ADS system has over 700,000 active users, with an average assisted driving mileage of nearly 700 kilometers per user in July alone [5][6] Group 2: Future Plans and Market Position - Huawei aims to accelerate the development of L3 and L4 autonomous driving technologies, targeting L3 commercial deployment on highways by 2026 and L4 pilot projects in urban areas by the same year [6] - The company has invested over 10 years in the automotive sector, with R&D expenditures exceeding 10 billion yuan in 2024 and a workforce of 8,000 dedicated to smart vehicle solutions [9] - Huawei's collaboration with automotive manufacturers focuses on empowering state-owned enterprises, enhancing their capabilities in electric and intelligent vehicle transitions [9]

Core Insights - As of July this year, the number of vehicles equipped with Huawei's QianKun intelligent driving system has reached 1 million, with over 10 million instances of parking assistance and a cumulative assisted driving mileage of 4 billion kilometers [1][3] - Huawei's QianKun intelligent driving system has successfully avoided potential collisions 2.54 million times, with an average assisted driving mileage of nearly 700 kilometers per person and an average of 37 parking assists per person [1][3] Industry Overview - The Chinese automotive assisted driving industry is experiencing intense market competition, with a significant reduction in the number of manufacturers compared to 5-6 years ago, indicating a trend towards market concentration [3] - The penetration rate of L2 level assisted driving functions in China was only 19% last year, but it is expected to soar to 50% this year, reflecting strong consumer demand for intelligent features [5] Technological Development - Huawei emphasizes the importance of data-driven technology in the development of assisted driving systems, arguing that the industry is fundamentally a competition of data volume, computing power, and algorithms [3][4] - The company has adopted a one-step end-to-end approach for its technology, which is seen as a more challenging but ultimately superior path compared to other manufacturers' visual language action models [3] Product Collaboration - Huawei has partnered with 28 vehicle models, including brands like Audi, Avita, and others, covering a wide range of vehicle types and price points, which has led to significant sales growth for its partners [5] - The collaboration with Audi signifies that luxury brands will increasingly consider intelligent features, presenting an opportunity for Chinese automakers to transition from mid-range to high-end markets [5] User Experience and Innovation - Huawei's MoLA architecture aims to integrate applications, hardware, and devices to enhance user experience, with features like a high-power audio system and AI assistants [6] - The company believes that a fragmented supply chain can lead to poor user experiences, thus it focuses on providing a cohesive system to ensure long-term satisfaction [6] Strategic Focus - Huawei's strategy is centered on long-term investment and user-centric development, with over 10 years of investment in its automotive business leading to current profitability [8][9] - The company prioritizes product quality and delivery in collaboration with partners, emphasizing that consumer perception of value is the key measure of success [9]